PSI - Issue 57

Giovanni M. Teixeira et al. / Procedia Structural Integrity 57 (2024) 670–691 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

685

16

exhaust manifold is shown in figure 15. Under real conditions the engine will stay heat up for much longer. However, this cannot (realistically) be tested. In contrast the thermal shock is performed in a much shorter time as the temperatures are forced down to room temperature in 720 seconds or less.

Fig. 14. Temperature distribution results from the Heat Transfer Analysis

In a thermal schock the ramping up and down of the temperature is non-linear (figure 15). Such profile can be achieved by adjusting the HTCs at every point in time or adjusting the time that matches the maximum tracked temperature.

Fig. 15. Temperature distribution results from the Heat Transfer Analysis

In the present case twenty time frames were used to describe the evolution of the temperature distribution along the 720 seconds of thermal shock. Therefore, twenty temperature field inputs were available to the subsequent structural analyses. The boundary conditions (for the stress analysis) are shown in figure 16. Nodes are constrained at the Z Direction (symmetry plane, blue) and at the X Direction (green plane) and Y Direction (bottom plane not shown). Exhaust manifold and cylinder head are connected by the bolts represented in red. The bolt load is 25kN. The gas pressure can be safely disregarded as its contribution to damage is rather small in comparison to the thermal load.